1 of 58

Working with Point Clouds in QGIS

Kurt Menke

Saber Razmjooei & Martin Dobias

2 of 58

Software

  • To get the most out of this workshop you are encouraged to install the nightly release of QGIS (v 3.43.0).
  • There are several important enhancements to working with point clouds which will be included in the next stable release of QGIS (v. 3.44), later in June. If you can run the nightly release you will have access to these important features.
  • Minimally, you will need the latest stable release - QGIS 3.42 Münster.

​

    • If using Windows, you can download it from the official QGIS website (https://qgis.org/download/) or by using the OSGeo4W installer.

​

    • For MacOS users, try to download and install the DMG file from this pull request (https://github.com/qgis/QGIS/pull/60039).

​

    • For Debian based users, you can use QGIS Flatpack or change the repo.

3 of 58

Resources

  • Step-by-step tutorial
  • Data package

4 of 58

What are Point Clouds?

  • A collection of points in 3D space.
  • Acquired through laser scanning (LiDAR) or photogrammetry.
  • The points are not uniformly distributed. Their density varies due to surface properties.
  • They have attributes including X | Y | Z coordinates.

5 of 58

What can you use Point Clouds for?

  • Data visualization
    • Classification of different types of objects.
    • Rendering by numeric attributes - including height (Z values).
    • 3D scenes.
    • Elevation profile plots.
  • Analysis
    • Measurements
    • Filter out specific objects.
    • Convert to both raster and vector data.

6 of 58

What types of files are used?

  • LAS – open binary format with attributes such XYZ coordinates, intensity, color and classification.
  • LAZ – Compressed version of LAS.

------------------------------------------------------

Storage Formats

  • EPT – Previous storage format - with indexing for quick access. Based on Entwine Point Tile (EPT).
  • COPC (Cloud optimized point cloud) – Current storage format. A variant of LAZ.

7 of 58

Point Cloud Support in QGIS

  • First added at QGIS 3.18 – spring 2021
  • New functionality added at each release since
  • The latest feature is the ability to edit point cloud attributes
    • Part of this is available in 3.42 but most is only in 3.43 (nightly release)

8 of 58

Loading Point Clouds in 2D

  • When you are loading a LAZ/LAS file in QGIS for the first time, QGIS indexes the file and creates a COPC file which will be used thereafter.
  • Depending on the size of your LAS/LAZ file, the initial indexing may take a while.
  • While the COPC is being generated you will see the Extent only on the Map Canvas as a red dashed outline

9 of 58

Loading Point Clouds in 2D

  • When a point cloud is loaded, QGIS follows a logic to select the best renderer:
    • if the dataset contains color information (red, green, blue attributes), the RGB renderer will be used,
    • else if the dataset contains a Classification attribute, the Classification renderer will be used,
    • else it will fall back to rendering based on Z attribute using the Attribute by Ramp renderer .

10 of 58

Loading Demo

11 of 58

Task 1 - Loading Point Clouds in 2D�

Pages 4-6

12 of 58

Point Cloud Properties

13 of 58

Point Cloud Attributes

14 of 58

Point Cloud Processing Tools�Information

15 of 58

Point Cloud Properties Demo

16 of 58

Task 2 - Point Cloud Properties�

Pages 6-8

17 of 58

Rendering Point Clouds in 2D

QGIS allows you to symbolize your point cloud data via four different renderers:

    • Extent only - Only a bounding box of the extent of the data is displayed
    • Attribute by Ramp - The data is stretched over a color gradient based on a numeric attribute such as Z value.
    • RGB - Render the data using red, green and blue color values.
    • Classification - The data is rendered using different colors for different classes.

​

18 of 58

Rendering Point Clouds Demo

19 of 58

Task 3 – Rendering Point Clouds in 2D�

Pages 9-13

20 of 58

Elevation Profiles Demo

21 of 58

Task 4 – Elevation Profiles �

Pages 14-16

22 of 58

Viewing Point Clouds in 3D Demo

23 of 58

Task 5 – Viewing Point Clouds in 3D�

Pages 16-20

24 of 58

Creating Virtual Point Clouds

  • Similar concept as virtual rasters.
  • Point cloud files are large and are typically delivered as tiles.
  • VPC files allow you to essentially merge a set of tiled point cloud files into a single layer while only creating a small JSON file.
  • It is a JSON file with a *.vpc extension.
  • It references a set of point cloud files (LAZ, LAS, COPC).
  • You can then run processing tools on the VPC (reproject, clip osv.).
  • Can be viewed in 3D environment.

25 of 58

Creating Virtual Point Clouds Steps

  • Load the 4 Hammerhus *.laz files
  • Remove the original *.laz files
  • Add the *.copc.laz files
  • Run the Build Virtual Point Cloud tool

​

Note: If you leave the optional parameters unchecked, the VPC file will be built very quickly as the algorithm will only read metadata of input files.

With any of the optional parameters set, the algorithm will read all points which can take more time.

​

26 of 58

Virtual Point Clouds (VPC)

​

​

​

27 of 58

Task 6 – Build Virtual Point Cloud

Pages 20-25

28 of 58

Using Point Cloud Processing Tools

  • Clip

​

  • Export to Raster

29 of 58

Using Point Cloud Processing Tools

30 of 58

Task 7 – Create a Raster DSM from a VPC

Pages 26-31

Note in version 3.42, you can only save to the VPC format, because the input layer is in the VPC format!

However, in 3.43.x it should be possible to save to COPC.

​

Also note that it is possible to enter a Filter Expression to limit the points in the output to a specific subset!

31 of 58

�Interpolate with TIN Challenge�page 31��Filtering Challenge�page 32��Calculate Density�page 33

32 of 58

Editing Point Clouds

  • The most recent functionality introduced is the ability to edit point cloud datasets.
  • The main editing functionality was introduced at QGIS v 3.42 and there will be additional editing features added at QGIS 3.44.
  • Editing takes place in the 3D View.

​

Note: editing is not available for Virtual Point Cloud layers or LAZ. This feature is only available for Cloud Optimized Point Clouds (COPC).

Additionally, it is only possible to edit the attributes, not location of points.

​

33 of 58

Editing Point Clouds Demo

34 of 58

Helpful Tips for Editing

  • Use a Point Symbol Style of Circle instead of Square. It can make it easier to see the points.

​

  • Under Layer Rendering set the Draw Order to Bottom to Top so the points are organized by their Z values. Note: If you have a large point cloud dataset, this setting can increase rendering time.

​

​

​

35 of 58

Helpful Tips for Editing…

  • Enable Eye Dome Lighting in your 3D view. This makes it easier to distinguish features.
  • You can also enable Eye Dome Lighting in your 2D View. To do this, open your Layer Styling Panel and select the Shading Renderer tab . Here you can enable Global Map Shading and Eye-Dome Lighting.

​

36 of 58

Helpful Tips for Editing…

  • It is very useful to have both your 2D and 3D view docked so you can work with both while editing.

​

37 of 58

Task 8 – Editing Classification Attributes

Pages 34-46

38 of 58

Using PDAL Wrench

  • PDAL is an acronym which stands for Point Data Abstraction Library.
  • PDAL is a C++ library for translating and manipulating point cloud data. It is very much like the GDAL library which handles raster and vector data.
  • GitHub repo with command descriptions:

https://github.com/PDAL/wrench

39 of 58

Using PDAL Wrench

  • You might have noticed every time you run a point cloud processing tool, it runs a PDAL wrench command. You could see that in the log tab of the dialog windows.

​

40 of 58

Task 9 – Using PDAL Wrench

Pages 46-54

41 of 58

Thanks!

​

​

Kurt Menke

​

kurt@septima.dk

Martin Dobias

​

martin.dobias@lutraconsulting.co.uk

42 of 58

Processing Tools

43 of 58

Interpolating Point Clouds

44 of 58

​

​

​

​

Interpolating Point Clouds

45 of 58

Filtering Point Clouds

​

​

​

​

46 of 58

Filtering Point Clouds via Toolbox

​

​

​

​

47 of 58

Filtered Point Cloud

​

​

​

​

48 of 58

Merging Point Cloud Datasets

49 of 58

50 of 58

Adding Point Clouds to QGIS

51 of 58

Layer Properties

​

​

​

52 of 58

Styling Point Clouds – 2D

​

​

​

53 of 58

Classification

​

​

​

54 of 58

Attribute by Ramp

​

​

​

55 of 58

RGB

​

​

​

56 of 58

Elevation Profiles

​

​

​

57 of 58

3D

58 of 58

New Book!

https://locatepress.com/book/mergin-maps

  • 132 pages
  • PDF $25

Coupon code 🡪 QGISUC24

20% all LocatePress e-books�

​